WhatsApp and Instagram, along with Facebook, seemed to have recovered from a historic internet outage that lasted nearly six hours — the longest ever time WhatsApp, Instagram or Facebook’s apps (including Messenger) and services have remained offline, inaccessible to billions of users around the world.
AFP
We are all used to the occasional outage suffered by WhatsApp or Instagram or any of these Facebook apps in isolation every now and then, but rarely have we witnessed the whole bouquet of Facebook services remain offline and inaccessible for such an unprecedented period of time.
In a weird twist of fate, the WhatsApp and Instagram outage happened just hours after a Facebook whistleblower revealed damning details of company culture where corporate profits take precedence over user safety and mental health — the revelations caused Facebook stock price to slide 5% and Mark Zuckerberg’s personal wealth decline by $7 billion or over Rs 50,000 crore in a single day.
Facebook has not yet revealed what caused the epic internet outage at WhatsApp, Instagram and Messenger, among other things, but it seems to be related to the July 2019 outage Facebook suffered due to inaccessible BGP routes into its network. Let me help and explain more plainly.
Jump To
What caused WhatsApp, Instagram, Facebook internet outage
Just like any other website, Facebook relies on external DNS servers (Domain Name System servers) for pointing online users to its domain’s IP address — think of it like an actual door number that your postal service has on its record to find and deliver physical mail addressed to you. Every website you visit online has its own unique IP address, and the DNS servers are like a postal service that convert www.facebook.com to the server record of Facebook’s IP address.
SOPA
ALSO READ: From Google To WhatsApp: Worst Server Outages In Recent Internet History
Without a working DNS, your computer’s browser or smartphone won’t know where to find Facebook or WhatsApp or Instagram, in effect stopping you from both posting updates and receiving updates from these services. Because Facebook is so huge — 2.85 billion monthly active users (separate to Instagram’s 1 billion users and WhatsApp’s 2 billion monthly active users) — it maintains a fairly complex set of DNS servers.
However, according to some solid initial reporting by Ars Technica, Facebook’s DNS servers weren’t wholly inaccessible during last night’s epic internet outage — since at least some of them are hosted on Amazon and they were functioning, but billions of requests around the world asking to connect to WhatsApp and Instagram and Facebook couldn’t be fulfilled because of BGP routing failure.
Enter the usual suspect — BGP failure
Cloudflare has a very good explanation on this, if you want to geek out a bit more, but for the uninitiated BGP stands for Border Gateway Protocol. It’s a system designed to exchange internet traffic routes that makes the whole internet work, a slightly advanced form of DNS, if you will. The entire BGP layer of the Internet has “constantly updated lists of the possible routes that can be used to deliver every network packet to their final destinations.”
If the DNS system is the postal service, then think of BGP as parallel courier networks that are constantly working with each other to help deliver mail from one place to another. Big tech companies (like Facebook, Google, Microsoft, etc) simply because of their enormous size and number of applications, services and user base maintain exhaustive BGP lists that are updated routinely to ensure users reach their services quickly without any issue.
ALSO READ: Why The Internet Keeps Going Down Frequently, Blocking Access To Your Favourite Websites
Think of Facebook as an entire walled off city (because it’s a private network) that needs to constantly announce its delivery routes to the country’s postal service through a shared system called BGP. Without BGP, the postal service won’t be able to find where Facebook’s city is located and deliver all the mail and traffic to all its apps and services. Without BGP, the Internet would simply stop working. And that’s sort of what happened with Facebook, WhatsApp and Instagram last night, because for most of us it’s kind of a large part of the Internet.
Reuters
While we still wait for official confirmation, Facebook appears to have goofed up its BGP routes, which prevented its DNS servers from finding all of its applications and services like WhatsApp and Instagram. And because BGP is a 30-year-old protocol, and because the internet is so vast and unbelievably complex in how it connects each and every one of us through cables and servers, updating BGP lists takes time to hit and register all corners of the Internet.
Regardless of the fact that BGP is old and imperfect, it’s the best system we have that makes the Internet work, and Facebook is still to blame for messing up a routine system update that resulted in one of the largest online outages in the history of the Internet.
Were you stressed about the Facebook, WhatsApp, Instagram outage last night or did it actually help you get a good night’s sleep? Let us know in the comments section below, and keep reading universo virtual.com for the latest science and technology news and insights.